Combine answers from all reasoning models(deepseek-r1, o1, flash-thinking, qwq)github.com/echohive421 pointechohive42a year ago